Examination of the aircraft revealed no deficiencies, no engine failure nor aircraft system failure. There is no indication of any emergency situation or aircraft problems prior to impact, and no distress calls were received. The pilot decided to go around after informing the pilot of the other aircraft that he had doubts regarding the suitability of the water landing area. The purpose of the manoeuvre was to fly a circuit and make another examination of the area so he could establish more clearly the characteristics and condition of the landing area. Consequently, the pilot's decision was justified from a safety standpoint, based on the information obtained during the approach. The pilot's decision to fly a left-hand circuit was sound since he was sitting in the left seat and visibility was better on that side. Since the reconnaissance phase was to allow an examination of the surface of the water, it required that the flight be made at low level and low speed. The reconnaissance track was to be rectangular, or oval. When the pilot turned onto the crosswind leg, he was over the river and, due to drift over the ground, the ground speed increased. Since the aircraft was nearly in level flight, it must have closed rapidly with the rising terrain. The pilot did not keep close to the north side of the valley before turning onto the crosswind leg, which did not allow him as much room as was possible, should he have needed it. As a result, not all the space available for the circuit was used to minimize the roll attitude of the seaplane in the turns and maximize the aircraft performance. The type of circuit chosen by the pilot reflected his intention to overfly the landing area at low altitude. Flying in mountainous terrain requires constant vigilance. Pilots must constantly compare their impressions with instrument readings. It could not be determined why the pilot continued flying at right angles to the ridge and did not try to avoid the mountain until the pilot of CF-RHI warned him. However, two possible explanations were identified to explain the pilot's delay in turning onto the downwind leg: the pilot may have been distracted and/or he may have been subjected to optical illusions. On one hand, the pilot may have been preoccupied with either planning the water landing or an untimely event in the cabin. It is possible that, after the go-around, most of his attention was devoted to continuously examining the landing area to his left and slightly behind him rather than to flying the circuit. Similarly, a distraction on board the aircraft, such as the passenger being indisposed by motion sickness, could have had the same result. On the other hand, while flying toward the rising slope, the pilot may have been subjected to optical illusions, which can be treacherous at low altitude and at near-stall speeds. After the message from the pilot of CF-RHI warning him of the imminent danger of collision with the mountain, the pilot of C-GIGK seemed to have responded, but he had little time to react and his room to manoeuvre may have been so limited that a half-turn could have led to a stall in a turn. An assessment of the situation may have led the pilot to decide to end the flight quickly and make a forced landing in the best possible conditions on the slope of the valley, as he had done at Lake Louise in 1994. The pilot apparently did not have enough time to switch off the electrical power and the fuel and prepare for a quick evacuation from the cabin. The information gathered and the statements recorded did not reveal the causes of the accident. Considering the light damage sustained by the cabin, the accident was survivable. However, the rapid onset and intensity of the post-crash fire gave the occupants no chance to evacuate the cabin in time. It could not be determined how the passenger's seat-belt was unfastened. It appears that no attempt was made to evacuate, since the pilot and passenger were found sitting in their seats and the pilot's door was still locked after the accident. Damage to the right wing was serious enough to cause a fuel leak, which, on coming into contact with an ignition source, started the fire. Engine heat and electricity were the two possible sources of ignition. Although witness statements and the damage caused by the fire suggest that the fire started on the right side of the cabin, and the most likely source of ignition was electrical, the investigation could not determine the exact source of ignition. However, engine heat would have been the only possible source of ignition if the master electrical switch was selected off prior to impact.Analysis Examination of the aircraft revealed no deficiencies, no engine failure nor aircraft system failure. There is no indication of any emergency situation or aircraft problems prior to impact, and no distress calls were received. The pilot decided to go around after informing the pilot of the other aircraft that he had doubts regarding the suitability of the water landing area. The purpose of the manoeuvre was to fly a circuit and make another examination of the area so he could establish more clearly the characteristics and condition of the landing area. Consequently, the pilot's decision was justified from a safety standpoint, based on the information obtained during the approach. The pilot's decision to fly a left-hand circuit was sound since he was sitting in the left seat and visibility was better on that side. Since the reconnaissance phase was to allow an examination of the surface of the water, it required that the flight be made at low level and low speed. The reconnaissance track was to be rectangular, or oval. When the pilot turned onto the crosswind leg, he was over the river and, due to drift over the ground, the ground speed increased. Since the aircraft was nearly in level flight, it must have closed rapidly with the rising terrain. The pilot did not keep close to the north side of the valley before turning onto the crosswind leg, which did not allow him as much room as was possible, should he have needed it. As a result, not all the space available for the circuit was used to minimize the roll attitude of the seaplane in the turns and maximize the aircraft performance. The type of circuit chosen by the pilot reflected his intention to overfly the landing area at low altitude. Flying in mountainous terrain requires constant vigilance. Pilots must constantly compare their impressions with instrument readings. It could not be determined why the pilot continued flying at right angles to the ridge and did not try to avoid the mountain until the pilot of CF-RHI warned him. However, two possible explanations were identified to explain the pilot's delay in turning onto the downwind leg: the pilot may have been distracted and/or he may have been subjected to optical illusions. On one hand, the pilot may have been preoccupied with either planning the water landing or an untimely event in the cabin. It is possible that, after the go-around, most of his attention was devoted to continuously examining the landing area to his left and slightly behind him rather than to flying the circuit. Similarly, a distraction on board the aircraft, such as the passenger being indisposed by motion sickness, could have had the same result. On the other hand, while flying toward the rising slope, the pilot may have been subjected to optical illusions, which can be treacherous at low altitude and at near-stall speeds. After the message from the pilot of CF-RHI warning him of the imminent danger of collision with the mountain, the pilot of C-GIGK seemed to have responded, but he had little time to react and his room to manoeuvre may have been so limited that a half-turn could have led to a stall in a turn. An assessment of the situation may have led the pilot to decide to end the flight quickly and make a forced landing in the best possible conditions on the slope of the valley, as he had done at Lake Louise in 1994. The pilot apparently did not have enough time to switch off the electrical power and the fuel and prepare for a quick evacuation from the cabin. The information gathered and the statements recorded did not reveal the causes of the accident. Considering the light damage sustained by the cabin, the accident was survivable. However, the rapid onset and intensity of the post-crash fire gave the occupants no chance to evacuate the cabin in time. It could not be determined how the passenger's seat-belt was unfastened. It appears that no attempt was made to evacuate, since the pilot and passenger were found sitting in their seats and the pilot's door was still locked after the accident. Damage to the right wing was serious enough to cause a fuel leak, which, on coming into contact with an ignition source, started the fire. Engine heat and electricity were the two possible sources of ignition. Although witness statements and the damage caused by the fire suggest that the fire started on the right side of the cabin, and the most likely source of ignition was electrical, the investigation could not determine the exact source of ignition. However, engine heat would have been the only possible source of ignition if the master electrical switch was selected off prior to impact. The pilot was certified, trained and qualified for the flight in accordance with existing regulations. Based on the autopsy and toxicology testing, there was no indication that incapacitation affected the pilot's performance. The weight and centre of gravity of the aircraft were within the prescribed limits. Records show that the aircraft was maintained in accordance with existing regulations. There was no indication of any airframe or flight controls failures, or of any engine malfunction. Weather conditions at the time of the accident were suitable for VFR flight, although the wind was moderate. After executing a go-around in the aux Mlzes River valley, the pilot turned left onto the crosswind leg to make a reconnaissance circuit about 450 feet above the river and 450 feet below the ridge of the valley. It seems that the pilot attempted to avoid the rising terrain after the other pilot warned him. The pilot died as a result of the post-crash fire, which started about five seconds after the accident. An unexplained distraction and/or the effects of an optical illusion may have contributed to distracting the pilot's attention from flying the circuit.Conclusions The pilot was certified, trained and qualified for the flight in accordance with existing regulations. Based on the autopsy and toxicology testing, there was no indication that incapacitation affected the pilot's performance. The weight and centre of gravity of the aircraft were within the prescribed limits. Records show that the aircraft was maintained in accordance with existing regulations. There was no indication of any airframe or flight controls failures, or of any engine malfunction. Weather conditions at the time of the accident were suitable for VFR flight, although the wind was moderate. After executing a go-around in the aux Mlzes River valley, the pilot turned left onto the crosswind leg to make a reconnaissance circuit about 450 feet above the river and 450 feet below the ridge of the valley. It seems that the pilot attempted to avoid the rising terrain after the other pilot warned him. The pilot died as a result of the post-crash fire, which started about five seconds after the accident. An unexplained distraction and/or the effects of an optical illusion may have contributed to distracting the pilot's attention from flying the circuit. The cause of the accident was not determined.Causes and Contributing Factors The cause of the accident was not determined. Although the cause of this accident was not determined, the conditions associated with it were conducive to optical illusions, resulting from flying over rising terrain at low altitude. To increase pilot awareness of this hazard, Transport Canada will be publishing an article on the subject in a future issue of its Aviation Safety Letter.Safety Action Although the cause of this accident was not determined, the conditions associated with it were conducive to optical illusions, resulting from flying over rising terrain at low altitude. To increase pilot awareness of this hazard, Transport Canada will be publishing an article on the subject in a future issue of its Aviation Safety Letter.